Molecular Rearrangements Involving Optically Active Radicals. X. The

Everett S. Wallis. The modem formulation of molecular rearrange- ments originally proposed by Jones and by Stieg- litz, and later developed by Whitmor...
0 downloads 0 Views 324KB Size
I (i74

JOHN

[CONTRIBUTION FROM

I?. LAYEA N D EVEKIITT S. WALLIS

THE

L'ol.

(;;j

FRICRCHEMICAL LABORATORY, PRINCETON UNIVERSITY]

Molecular Rearrangements Involving Optically Active Radicals. X. The Retention of Configuration in the Wolff Rearrangement BY JOHN F. LANE'AND EVERETT S. WALLIS The modem formulation of molecular rearrangements originally proposed by Jones and by Stieglitz, and later developed by Whitmore2 gives primary emphasis to changes occurring in a reactive intermediate designated : A : B, where A and B are atoms such as carbon or nitrogen. The rearrangement process takes place with an internal oxidation and reduction in :A :B which leads to the transfer from A to B of a pair of electrons together with the atom or group held by it. For a consideration of this transfer when optically active groups are involved, it is convenient to follow the ideas expressed in Part I of this seriesa and to write

R:

R,

This formulation is based on the assumption that the rearranging system a t no time dissociates into neutral or charged free fragments, Le., the transformation is strictly intramolecular. Under these conditions the optically active group should not racemize during the process of rearrangement. Moreover, the formulation admits no mechanism for the inversion of configuration.* Previous parts of this series3s5have shown that the Curtius, Hofmann and Lossen degradation^^^^"^^^" of opti(1) Charlotte Elizabeth Procter Fellow in Chemistry, 1940-1941. (2) For a discussion of the development of this formulation and its original applications see Whitmore, THISJOURNAL,64, 3274 (1932). (3) Jones and Wallis, ibid., HI, 169 (1926). (4) I t should be noted that this formulation is concerned primarily with the behavior of a group transferred with its electron pair from A to B. For experimental work on the course of rearrangement where A or B itself is an optically active group, cf, McKenzie, Roger and Wills, J . Chem. Sac., 779 (1926); McKenzie, Roger and McKay, ibid., 2597 (1932); Roger and McKay, ibid., 332 (1933); Wallis and Bowman, J . Org. Chem., 1, 383 (1936); Bernstein and Whitmore, THISJOURNAL, 61, 1324 (1939). Here, in accordance with modern principles of aliphatic substitution [Olson and Long, ibid., 66, 1294 (1934); Bartlett and Packel, ibid., 69, 820 (1937); Hughes, Trans. Faraday Sac., 84,202 (1938); Criegee and Plate, Bcr., 78, 178 (1939) 1, inversion of configuration accompanied by varying amounts of racemization appears t o be the rule. 61, 2892 (1929); (b) Wallis and (6) (a) Wallis, T H I S JOURNAL, Nagel, i b i d . , 63, 2787 (1931); (c) Wallis and Dripps, ibid., 66, 1701 (1933); (d) Wallis and Moyer, ibid., 2598 (1933); (e) Sprung and Wallis, ibid., 66, 1716 (1934); (f) Gilbert and Wallis, J . Org, Chem., 6, 184 (1940); (9) Lane and Wallis, ibid., 6, No. 3 (1911); (h) Lane, Willenr. Weissberger and Wallis, i b i d . , 6, 276 (1940).

cally active acid azides, amides, and hydroxamic acids, and the Wolff rearrangement of optically active diazoketones6g'hcontaining a quaternary asymmetric carbon atom proceed without appreciable racemization of the migrating group and are true intramolecular p r o ~ e s s e s . ~ ~It, ~has ,~ also been shown that the rearrangement of optically active alkyl phenyl ethers which leads to optically active products cannot be explained on the basis of an intermolecular mechanism.6f Until now the difficulty of relating rotation to configuration has prevented a rigid confirmation of the absence of the Walden inversion which is predicted by I. It is true that Bartlett and KnoxG have shown that in hindered systems the Hofmann rearrangement may occur without inversion, and that a similar result has been obtained by Noyes7 in the asymmetric degradation of the camphoric acids. But in the absence of a rigorous configurational relationship in the system RIRzCHCOOH+ R I R ~ C H N Ha~ ,general proof that this process is accompanied by retention of configuration cannot be constructed. This important prediction of the formulation (I) may, however, be rigorously tested by means of the Wolff rearrangement of an optically active diazoketone, for the necessary configurational relationships are readily obtainable. We have achieved this by the cyclic process shown. The acid chloride of (+)-2-methyl-2-phenylhexanoic acid, [a]2 0 +24.0° ~ (benzene), was converted by the action of diazomethane to (+)-1diazo - 3 - methyl - 3 - phenylheptan - 2 - one, [a]2 0 +65.0° (benzene). This compound was then allowed to rearrange in a solution of aqueous dioxane containing sodium thiosulfate and silver oxide. The (-)-3-methyl-3-phenylheptanoic acid, -6.35' (1-dm. tube without solvent), so formed was converted by diazomethane to its methyl ester, [ a ] " ~-114.9' (benzene). This ester was subjected to a Barbier-Wieland degradation. It gave (+)-2-methyl-2-phenylhexanoic acid. Thus, i t was found to be configurationdlly identi(6) Bartlett and Knox, THIS JOURNAL, 61, 3184 (1939). (7) Noyes, A m . Chem. J . , 16, 500 (1894); Noyes and Potter, THIS JOURNAL, 37, 189 (1916); 34, 1067 (1912); Noyes and Nickell, ibid , 56, 118 (1914).

~

in. p. 88" [a]*'D +29"

f+\R-CO,H

Barbier-Wieland degradation

(-)R-CH~-COZCHI

(4)

/

CH3 \ R = n-CdH9-C-

i j

CeHs

cal with the original diazoketone. The acid so obtained was identified as the crystalline (+)-!2methy 1- 2-phenylhexano -$-bromo anilide, m .p . SSO , [a] ?OD +29.0° (benzene), a derivative identical in every respect with that prepared similarly from the original (+)-2-methyl-2-phenylhexanoicacid from which the diazoketone was prepared. Since only step (2) in the cycle (11) involved the breaking of a bond of the asymmetric center, it is clearly established that in this step the Wolff rearrangement proceeded without a Walden inversion, This is in accordance with the prediction of (I), We therefore submit that in any intramolecular rearrangement, a group migrating with ita electron pair does not suffer a Walden inversion. It should be noted that the only possible objection to be raised against this generalization is that the racemizations observed in Wolff rearrangements of (+)-benzylmethyldiazoaceto~ie,~~ indicate a real difference between the rearrangement mechanism for this process and for other processes assumed to be of type (I), i.e., the Curtius rearrangement of (+)-benzylmethylacetazide where no racemization is ~ b s e r v e d . ~We have presented experimental evidence against this objection elsewhere6g and have attributed the observed racemizations to enolization of the intermediate //O R~R~CH-C-CH prior to rearrangement, the enolization being markedly catalyzed by the catalytic surfaces Ag', Ag(s) or Ag', ,4g(s), AgzO(s) which are necessarily present if the rearrangement is to be conducted at temperatures below 180'. It would be of great interest in this connection to conduct a quantitative investigation of the influences of such catalytic surfaces on the rate of enolization of the optically active ketone, (CsH6C~-I~)(CH~)CHCOCGHS, and to compare the results obtained with those of Conant and CarlsorP on the homogeneous rate of enolization of this [\)

C mi t i i t .end Cdrlson T H IJ~O L R ~ A I 54, W i . 7 (193?)

compound. Such an investigation is contemplated in the near future.

Experimental Preparation of ( +)-2-Methyl-Z-phenylhexanoic Acid.One hundred ten grams of this acid was prepared and resolved by the method of Conant and Carlson.x After three recrystallizations from dilute alcohol, the quinine salt of the acid had a specific rotation [ a ] " D -87.0' ( c = 2.60, chloroform). Further recrystallization did not alter this value, The quinine salt on decomposition with dilute hydrochloric acid gave 27 g. of an acid of specific rotation [a]''gse,~ f18.4" [aIZ051~1-24.0"; [a]"54es +29.0° (c = 3.80, benzene); [ a ] 2 0 5 4 6 3 +14.3' (c= 3.80, alcohol) (Conant and Carlson report +13.7" under these conditions). For identification, the acid was converted to the corresponding p-bromoanilide by the action of p-bromoaniline on the acid chloride. Three recrystallizations from dilute alcohol gave a product which melted at 88" and had the following specific rotations: [ a l Z o 8 s e 3 +22.1"; [ a l z o 5 ~ ~ +28.9'; [ a ] 2 0 6 4 8 f34.3" ~ (c = 2.50, benzene). Calcd. for C19H220NBr: C, 63.33; H, 6.16; N, 3.89; Br, 22.18. Found: C, 63.3; H, 6.1; N, 3.7; Br, 22.4. Preparation of ( - )-3-Methyl-j-phenylheptanoic Acid.By a method previously described,6g10 g. of (+)-2-methyl2-phenylhexanoic acid was converted to (+)-1-diazo-3methyl-3-phenylheptan-2-one,[ a ] , ' ~f65.0" (c = 14.1, benzene). This compound was then subjected to the Wolff rearrangement in a solution of aqueous dioxane containing silver oxide and sodium thios~lfate.~' The product, ( -)-3-methyl-3-phenylheptanoicacid, boiled a t 100" (0.01 mm.): a2h83 -4.89"; a Z 0 5 ~ ~ 3-6.35'; ~ L S S -7.70' (1-dm. tube without solvent). For identiiication, a portion of the acid was converted to the corresponding anilide by the action of aniline on its chloride. Two recrystallizations of the crude material from dilute alcohol gave the pure anilide: m. p. 76-77'; mixed m. p. with an authentic specimenjPof (-)-3-methyl[a]206688 -47.0' ; 3-phenylheptanoanilide, 76-77' ; -60.2'; [a]zo,, -73.5" (c = 1.20, benzene). Preparation of ( -) -3-Methyl-3-phenylheptanoic Acid Methyl Ester.-To a solution of 2 g. of (-)-3-methyl-3phenylheptanoic acid in 25 cc. of ether a t 0" was added slowly and with shaking a solution of 1.0 g. of diazomethane in 50 cc. of ether. The mixture was allowed to stand for several hours at room temperature. I t was then filtered, and the cxce.;? of tiiazoinethane and ethcr x a s rcmovrd on

1676

JOHN

F. LANEAND EVERETT S. WALLIS

VOl. 63

the steam-bath under reduced pressure. The methyl ester removed under reduced pressure. The residue was diluted S S [ O ~ ] ~ ~ S S S S with 100 cc. of 2 N hydrochloric acid and extracted with boiled at 102-103" (2 mm.): [ O ~ ] ~ ~ ~ S-11.2"; -14.9'; [ O ~ ] ~ O W ~-17.4' S (c = 4.00, benzene). ether. The ether solution was washed k s t with 2 N hydroBarbier-Wielandg Degradation of (-)-3-Methyl-3chloric acid and then with water. The organic acid was phenylheptanoic Acid Methyl Ester.-The methyl ester then extracted from it with 5% potassium hydroxide solu(1.8 9.) dissolved in 15 cc. of dry benzene was added tion. The alkaline extract was washed with ether and slowly with shaking in the cold t o a Grignard reagent pre- acidified with hydrochloric acid. The process of extraction pared from 20 g. of bromobenzene and 3.2 g. of finely cut with ether, washing, and final extraction with 5% potasmagnesium ribbon in 50 cc. of anhydrous ether. The mix- sium hydroxide solution was repeated. The alkaline exture was refluxed for two hours on the steam-bath. Twen- tract was acidified with hydrochloric acid, and the organic ty-five cubic centimeters of dry benzene was then added, acid was taken up in benzene. The benzene solution was and the solvent was distilled as completely as possible over washed with water, atered, and evaporated under reduced a period of two hours. The residue was heated for an pressure. One-half gram of an acid was obtained which additional two hours on the steam-bath. It was then had a specific rotation [al2Oo+20.0' (c = 2.50, benzene). cooled, diluted with benzene, and poured over a mixture Identification of the Acid Obtained in the Barbierof crushed ice and ammonium chloride. The benzene solu- Wieland Degradation.-By the action of p-bromoaniline tion of the product was filtered and evaporated. The upon its chloride, the acid obtained as described in the preresidue was then refluxed for two hours with 30 cc. of 0.5 N ceding paragraph was converted t o the corresponding palcoholic sodium hydroxide t o saponify any unchanged bromoanilide. The crude material had a specific rotation ester. Most of the alcohol was removed under reduced [a]*%+23.0' (c = 2.75, benzene). Two recrystallizapressure, and the residue was dissolved in benzene, washed tions from dilute alcohol gave pure (+)-3-methyl-3-phenylseveral times with water, atered, and evaporated. The heptano-p-bromoanilide: m.p. 87-88'; mixed m. p. with residue was then distilled with steam for six hours to re- the analytical sample described in the first paragraph, 87move diphenyl. Finally, the product was taken up in 88"; [ a J Z o 6 ) 8 ~ f22.0'; [a]206ps3 $29.0'; [a.lZo~ss benzene and the iiltered solution evaporated under reduced $34.0" (c = 1.00, benzene). pressure. The crude ( -)-1,1,3-triphenyl-3-methylheptanThe authors are indebted to Merck and Com1-01 (2.6 9.) so obtained had a specific rotation [ 0 1 ] ~ o D pany, Inc., Rahway, New Jersey, for the analyses -16.3' (c = 3.00, benzene). published in this paper. The carbinol was then refluxed for thirty minutes with 10 cc. of acetic anhydride. After the excess of the reagent Summary had been removed under reduced pressure, the residue was (-)-3-Methyl-3-phenylheptanoic acid has been treated with hot water to destroy traces of the anhydride. The product was taken up in benzene, washed with water, prepared by a Wolff rearrangement of (+)-1and the solution filtered and evaporated under reduced diazo-3-methyl-3-phenylheptan-2-one. A Barpressure. The crude ( -)-1,1,3-triphenyl-3-methylhep- bier-Wieland degradation of the methyl ester of tene-1 so obtained had a specific rotation [cy]% -20.2' this acid gives (+)-2-methyl-2-phenylhexanoic (c = 2.50, benzene). Without further purification, the olefin was dissolved in acid, which is configurationallyidentical with (+)1 diazo 3 methyl 3 phenylheptan 2 one. 20 cc. of purXed glacial acetic acid (distilled over potassium permanganate). To this solution at 15' was added The Wolff rearrangement, therefore, proceeds with mechanical stirring a solution of 2.5 g. of chromium without configurational inversion of the migrating trioxide in 2 cc. of water and 5 cc. of p d e d glacial acetic acid. The addition was regulated so as to consume one- group. A discussion of these results is given in terms of half hour, and stirring was continued at 15' for an additional fifteen minutes. The solution was then chilled to the modern electronic concept of molecular re-loo, and the excess of chromium trioxide was decom- arrangements. It is concluded that in any intraposed at such a rate that the temperature remained below molecular rearrangement, a group migrating with 10". When this had been accomplished, the solvent was

-

(9) Barbier and Locquin, Comfit. rsnd., 186, 1443 (1913); Wieland. Schlichting and Jacobi, Z. fihysiol. Chem.. 161, 80 (1926); cf. also Hoehn and Mason, THISJOURNAL, 80, 1493 (1938).

- -

- -

- -

its electron pair does not suffer a Walden inversion. PRINCETON, NEW JERSEY RECEIVEDAPRIL3,1941